AI economics: Revolut ditches Claude Code for self-hosted GLM
Revolut says it cut AI coding costs about 10x by making a self-hosted Chinese GLM model its default. Separate commentary raised model-collapse worries and AI regulation motives.
What we know
Revolut runs Z.ai's GLM on its own Nvidia GPUs at about $0.30 per million tokens. An account on X amplified an Oxford model-collapse claim, while McKenna argued AI labs push regulation to protect their moat.
By the Global Eye 24 desk, with AI assistance. Updated as the story develops. Spot an error? Write to corrections@globaleye24.com.